[fix](test) fix regression test eliminate_inner

[feiniaofeiafei]

What problem does this PR solve?

Issue Number: close #xxx

Related PR: #xxx

Problem Summary:

Release note

None

Check List (For Author)

• Test

  • Regression test
  • Unit Test
  • Manual test (add detailed scripts or steps below)
  • No need to test or manual test. Explain why:
    • This is a refactor/code format and no logic has been changed.
    • Previous test can cover this change.
    • No code files have been changed.
    • Other reason

• Behavior changed:

  • No.
  • Yes.

• Does this need documentation?

  • No.
  • Yes.

Check List (For Reviewer who merge this PR)

• Confirm the release note
• Confirm test cases
• Confirm document
• Add branch pick label

[hello-stephen]

Thank you for your contribution to Apache Doris.
Don't know what should be done next? See How to process your PR.

Please clearly describe your PR:

1. What problem was fixed (it's best to include specific error reporting information). How it was fixed.
2. Which behaviors were modified. What was the previous behavior, what is it now, why was it modified, and what possible impacts might there be.
3. What features were added. Why was this function added?
4. Which code was refactored and why was this part of the code refactored?
5. Which functions were optimized and what is the difference before and after the optimization?

[feiniaofeiafei]

run buildall

[feiniaofeiafei]

run nonConcurrent

[feiniaofeiafei]

run nonConcurrent

[feiniaofeiafei]

run buildall

[morrySnow]

/review

[github-actions]

Automated review completed for PR 64457.

No blocking issues found. The change is a narrow regression-test stabilization: eliminate_inner is a PK-FK inner-join elimination shape suite, and INFER_SET_OPERATOR_DISTINCT is an orthogonal set-operation rewrite that changes the shape of the suite's UNION DISTINCT inputs without changing the result assertions. Disabling it here keeps the suite focused on the join-elimination rule being tested.

Checkpoint conclusions:

• Goal and proof: The PR aims to stabilize/fix the eliminate_inner regression test. The touched suite still checks both explain shape plan and ordered query results; GitHub shows P0 Regression succeeded for this head SHA.
• Scope: The modification is minimal and focused: one session setting in one regression suite.
• Concurrency, lifecycle, locks: Not applicable; this is test setup only.
• Configuration/compatibility: No new config or protocol/storage compatibility change. It uses an existing session variable and existing rule name.
• Parallel code paths: No runtime path change. The disabled rule has its own set-operation coverage elsewhere; this suite remains scoped to PK-FK join behavior.
• Test coverage/results: No .out update is needed because the expected shape/result file remains unchanged under the new setting; CI P0 regression passed.
• Observability, transaction/persistence, data writes, FE-BE variable passing, performance: Not applicable to this test-only change.
• User focus: No additional user-provided review focus was present.